[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-121110-en":3,"doc-seo-121110-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},121110,8796095461610,"Oliver","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",6,"Technology","A Machine Learning Pipeline for Automated Insect Monitoring","Climate change and anthropogenic pressures have driven severe declines in insect populations, threatening biodiversity and ecosystem services. Reliable insect abundance data remains insufficient for scalable ecological assessment. This work presents an open-source, end-to-end machine learning software pipeline for automated moth monitoring using camera traps, covering insect localization, moth/non-moth classification, fine-grained moth species identification, and individual tracking across frames, enabling massively scalable data collection for entomology.","A machine learning pipeline for automated insect monitoring  \narXiv :2406 . 13031v1 [ cs .CV] 18 Jun 2024  \nAditya Jain∗† Mila-Quebec AI Institute  \nFagner Cunha∗ Michael Bunsen∗  \nFederal University of Amazonas Mila-Quebec AI Institute  \nLéonard Pasi‡ EPFL  \nAnna Viklund‡ Daresay  \nMaxim Larrivée  \nMontreal Insectarium  \nDavid Rolnick  \nMcGill University Mila – Quebec AI Institute  \nAbstract  \nClimate change and other anthropogenic factors have led to a catastrophic decline in insects, endangering both biodiversity and the ecosystem services on which human society depends. Data on insect abundance, however, remains woefully inadequate.  \nCamera traps, conventionally used for monitoring terrestrial vertebrates, are now being modified for insects, especially moths. We describe a complete, open-source machine learning-based software pipeline for automated monitoring of moths via camera traps, including object detection, moth/non-moth classification, fine-grained identification of moth species, and tracking individuals. We believe that our tools, which are already in use across three continents, represent the future of massively scalable data collection in entomology.  \n1 Introduction  \nThe Earth is undergoing a sixth mass extinction event, where an eighth of all species may become extinct by 2100 [1–3] . Insects account for about half of all living species on earth and 40% of the animal biomass [4], but both the diversity and abundance of insects are undergoing a precipitous decline [5] as a result of several factors, in which climate change figures prominently. The “insect apocalypse” significantly increases the risk of breakdown of ecosystem functions on which human society depends [6] . Monitoring insects is therefore a crucial component of climate change adaptation.  \nTraditional insect collection and identification by entomologists is hard to scale, due to the massive number of insect species and a lack of experts, with certain geographies and taxonomic groups especially poorly covered. The emergence of high-resolution cameras, low-cost sensors, and processing methods based on machine learning (ML) has the potential to fundamentally change insect monitoring methods [7] . Camera traps powered by computer vision models for terrestrial vertebrates monitoring are now commonplace [8], and specialized camera trap hardware for insect monitoring has begun to gain momentum [9–13] . A common group of focus for such studies has been moths[14–16], which serve vital ecological roles and represent a fifth of all insect species. Importantly, most moths can readily be attracted with UV light and are frequently visually distinguishable up to species or genus, making them ideal targets for camera traps. As hardware for moth-monitoring has grown more common, however, there is a need for scalable data processing techniques to match the influx of data. Prior methods have been greatly limited in the species and geography covered, as well as requiring extensive manual labelling for training the algorithms.  \n∗Equal contribution.  \n†[Correspondence to:](Correspondence to: moth-ai@mila.quebec)[ moth-ai@mila.quebec](Correspondence to: moth-ai@mila.quebec)[ ](Correspondence to: moth-ai@mila.quebec)‡Work done while at Mila-Quebec AI Institute.  \nTackling Climate Change with Machine Learning: workshop at NeurIPS 2023 .  \n(a) A moth camera trap (b) Trap in operation (c) Machine learning predictions on a raw image  \nFigure 1: Figure 1a and Figure 1b depict a moth camera trap used by our partners. Figure 1c shows our insect localization and species prediction on a raw image, with red boxes showing insects classified as moths (with fine-grained species predictions), and blue boxes showing non-moths.  \nThis work describes a complete software and ML framework that transforms raw photos from insect camera traps into species-level moth data (see Fig. 1) . Our system is motivated by the following goals:  \n1) Model predictions should be highly accurate across mo","cbCaih6a0MtxBvis","https://ap.wps.com/l/cbCaih6a0MtxBvis","pdf",10393499,1,9,"English","en",105,"# Introduction\n# Machine learning pipeline\n## Training data for image classification","[{\"question\":\"What problem does the machine learning pipeline address?\",\"answer\":\"It addresses the shortage of scalable insect abundance data by automating moth monitoring from camera trap images and reducing reliance on extensive manual labeling.\"},{\"question\":\"How does the pipeline process camera trap images?\",\"answer\":\"It uses a multi-stage workflow: object detection for insect localization, binary classification for moth vs non-moth, fine-grained moth species classification, and tracking across frames to count individuals.\"},{\"question\":\"Where does the pipeline get training data without labeling trap images directly?\",\"answer\":\"It uses zero-shot transfer learning with annotated data from GBIF, especially iNaturalist, combined with unlabeled moth trap data provided by partner ecologists.\"}]","A Machine Learning Pipeline for Automated Insect Monitoring | PDF",1785733781,23,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"a-machine-learning-pipeline-for-automated-insect-monitoring","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/technology/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/a-machine-learning-pipeline-for-automated-insect-monitoring/121110/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-03",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does the machine learning pipeline address?","Question",{"text":75,"@type":76},"It addresses the shortage of scalable insect abundance data by automating moth monitoring from camera trap images and reducing reliance on extensive manual labeling.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the pipeline process camera trap images?",{"text":80,"@type":76},"It uses a multi-stage workflow: object detection for insect localization, binary classification for moth vs non-moth, fine-grained moth species classification, and tracking across frames to count individuals.",{"name":82,"@type":73,"acceptedAnswer":83},"Where does the pipeline get training data without labeling trap images directly?",{"text":84,"@type":76},"It uses zero-shot transfer learning with annotated data from GBIF, especially iNaturalist, combined with unlabeled moth trap data provided by partner ecologists.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,113,118,123,127,130,134],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":111,"slug":112},50,"technology",{"id":114,"doc_module":4,"doc_module_name":46,"category_name":115,"show_sort_weight":116,"slug":117},7,"Healthcare",40,"healthcare",{"id":119,"doc_module":4,"doc_module_name":46,"category_name":120,"show_sort_weight":121,"slug":122},8,"Research & Report",30,"research-report",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]